Fracesco Cetraro of .Cloud provides these reflections for a new user-centric approach to getting people online.
There seems to be general consensus (and concern) that new TLDs have not made the impact expected and that overall adoption is lagging behind. However opinions are definitely more divided regarding the causes of the problem and the possible solutions.
Some believe it is ICANN’s fault for not doing enough to advertise new extensions to the general public in spite of having significant resources to do so. Others blame Registrars for lack of effort or the Registries for having unrealistic expectations. Some also blame the end users for either not understanding the value of new TLDs, or for simply not caring at all.
The situation is rendered even more paradoxical by the fact that statistics continue to show that a large number of small and medium businesses even in the more advanced economies still do not have an online presence, yet the large majority of companies in the domain and hosting business are growing very slow, if often at all.
Reflecting over these issues, one thing that strikes me as odd is nobody seems to be questioning the mindset with which we approach the task of helping our customers to get online.
I think few will disagree with the observation that over the past 20 years the type of customers we deal with has changed significantly. “Techies” have been increasingly replaced by a much broader and diversified consumer base that does not care much about gigabytes and databases and “just wants to get online” quickly, at the right price and without hassles.
In the same time period, we have added new ”innovative” products to the mix, we started using Social Media for marketing and made our shopping carts responsive. However the approach used to sell our services – a reflection of the underlying assumptions we make almost unconsciously on ”how domain/hosting products should be sold” – has basically never changed. The typical sales flow still starts by picking a domain, then a hosting package, followed by various add-ons and then eventually collecting payment.
The philosophical problem with this approach is that it seems to completely disregard Hick’s Law and instead take for granted that in a world of exponentially growing choices, the average customer will know exactly what he/she wants and is able to make the best decision.
Let me try to illustrate the issue with a “restaurant” example. Imagine you have never had Thai food, but a friend tells you it’s great and you decide to try it. You enter a crowded restaurant and are unceremoniously handed a thick menu with hundreds of possible dishes to choose from. Overwhelmed, you order the one dish you have heard about and hope for the best. Sure, you are most likely going to get full, but are you really getting the best experience and value?
Imagine now the same scene, but in a restaurant where the staff welcomes you and instead of the thick menu gives you some general indication about Thai cuisine, asks you about your preferences and recommends a set of dishes that is most likely to satisfy you and give you an amazing experience. I bet you are going to leave this second restaurant a lot more satisfied. You’ll tell your friends about what a great experience you had and you are most likely going to go back each time you have cravings for Thai food (also, I hope you tipped that waiter generously :)).
Today, the average customer experience when buying domains and hosting is more likely to resemble the first example than the second. I doubt anyone enjoys dealing with a thick, unfamiliar “menu”, regardless of whether it is Thai delicacies or lists of domains, site-builders or email solutions, without any indication of what is ”right for me”. Expecting the average customer to make the best decision in these conditions is ”naive” at best and extremely unlikely to work (so no wonder it’s not working…).
Nothing exemplifies this paradox better than the issue of ”domain search”. Many of the people I talk to seem to believe that if we only manage to ”fix search” then all our problems will be solved and new TLDs’ adoption will finally take off.
But consider for a second the challenge that building an “efficient” domain search function faces. Particularly with the advent of new domain extensions and the exponential increase in possible combinations. The complexity that domain search has to deal with is just too big. No matter how much effort we put into “fixing” it, how can we reasonably expect it to provide relevant results when all we feed into it is a couple of keywords that the user types in the search bar? I am sure that even the smartest Google engineer would be gasping for air if all he/she had to work with was a couple of vague keywords while still being expected to deliver the most relevant ad for that user.
I am strongly convinced that the real issue facing new TLD adoption is not lack of demand, but the fact that we need to re-learn how to sell them in a way that makes sense for the user they are intended for. As long as the customer is presented with a “thick menu” it is very hard to blame him for going with the “safest” choice, instead of considering the possibility that there might be a better option for him.
My strong belief is that the value of new TLD is not in the simple fact that they are new, but rather in the fact that they offer relevance and a better way to add useful context to an online presence. However, how can we really provide the ”right” domain to the ”right” customer when we ourselves have no context to work with?
People don’t buy domains and hosting because it’s cool. They buy them because those are the instruments to build an online presence that can help them pursue their ultimate goals, whether these are promoting a business or sharing their opinions on a blog.
Starting a user interaction with the domain search is the equivalent of asking customers straight away ”what do you want”, which for the large majority is a very difficult question to answer well enough. A better approach would probably be to ask ”who are you and where do you want to go”, and use the information to guide them through the process of picking the right tools to get them there.
A better understanding of their “purpose” is key to evolving what is currently an obsolete, price-obsessed, one-size-fits-all, supermarket style of selling domains and hosting. It would also make it easier for the customers to understand the true value of things like backing up their data to the cloud for increased security, how having a well-performing website affects their ROI, or how a fitting domain can help their business be more visible. Once there is a clear perception of the value, it is also easier to move away from a race to the bottom that is only focused on competing with the lowest price.
The good news is that the data and the tools are there, we just need to learn to use them properly. Many players in this space are in my opinion moving in the right direction but no one has yet fully connected all the dots. To do so, it would require a fresh look at how we do things, for instance by daring to rip apart sales flows that haven’t changed in 20 years and find new ways to serve each specific customer on his own terms, and help choose the right tools to fulfill their goals.
Instead of forcing them to pick a domain first, we could let them build their site first, so that a smarter domain assistant can analyze their content to provide them with relevant domain suggestions and other services.
A happy customer whose business is growing tends not only to stick around longer, but also to spend more and more over time. While domains and hosting are already commodities, knowledge and ability to provide the right piece of advice are not: only those players that will be able to transition from “dumb suppliers” to “trusted advisors” will remain relevant, competitive and thrive.
@Francesco Cetraro,
There’s room for improvement in the domain search algorithm itself, though, as you acknowledge, it’s a complex, sometimes almost futile task to infer relevant domain names from a keyword query.
Prior to the nTLD rollout, I teamed up with another guy (a software engineer outside the domain industry) to design a new search algorithm for registrars and market places. But the domain industry is so little understood by outsiders that we couldn’t get a hearing. And, without backing, the project fizzled out.
I agree with you that the scope of the problem goes beyond domain search functionality. Scant input. Relevant to whom for what? We need to understand the person, the project, the audience. Which is why consultants like me interview clients before advising them which domain names to consider.
Ideally the sales funnel at registrars would be a slower process of interviews, brainstorming, consultation. However, most consumers want what they want; and they want it now. Only a few are willing to pay to hear that their first idea may not be the best idea.
A slower approach to selling domains might not be cost effective for registrars; and, with a greater chance of cart abandonment, it may backfire.
Also, I’d question whether registrars could ever truly be “trusted advisors”. Their natural incentive is to up-sell. And registrars are susceptible to registry kickbacks. Both tendencies are a conflict of interest. Any trusted advisor must be not only expert but neutral.
But registrars could definitely transition from “dumb suppliers” to smart suppliers. The best idea in your article, I think, is the penultimate paragraph:
“Instead of forcing them to pick a domain first, we could let them build their site first, so that a smarter domain assistant can analyze their content to provide them with relevant domain suggestions and other services.”
Some code-free web design services do try that. Hopefully they’ll succeed. We need more people building an online presence more easily.
Thanks Joseph!
TBH I don’t think every Registrar/Hoster has the potential and the mindset to become a “trusted advisor”. However, the fact that many are unable to adapt doesn’t make the customer’s need for it disappear.
I think we all agree that what we are doing is not working (or at the very least, has a lot of room for improvement), which is why I think it makes sense to consider the possibility there might be a different/better way of solving this problem.
While I don’t think I personally have the answers, I do have a lot of questions and I see an opportunity for someone to come in and take a whole new approach and potentially completely disrupt this space, maybe with the use of AI that help bring a better result without slowing down or disrupting the customer experience.
Rightside has put a lot of effort into domain search. I believe their new system is active on Name.com and they’re offering it to other registrars.
Rightside certainly did a good job with their domain search, but they are not the only ones who have been working hard on this. TBH, I think also Gandi did a terrific job with theirs.
Andrew, this is a brilliant and incredibly helpful post. Really well done. I am sharing internally and externally.
Thanks for writing this!
Sean
Annnd… just saw the byline. Thanks Francesco!.
wait…wait…let me hide that byline now that you called it brilliant 🙂
hehehe… too late 😉
Thanks Sean, I am happy you enjoyed it.
I blame the destruction of the functionality of the URL bar in browsers. Google, in other words.
Francesco,
This is a great post and I think it definitely touches on one of the most significant challenges facing our industry. I’ve written a few posts that capture the same ideas that you’re presenting here.
http://rightside.news/new-tlds-opportunity-challenge-part-1
http://rightside.news/new-tlds-opportunity-challenge-part-2
http://rightside.news/the-state-of-search
With the launch of nTLDs we’ve transitioned from roughly 10% of TLDs being ‘vertical’ to around 60% having a specific meaning. This creates a large opportunity, but also a big challenge. The approach that Registrars have used in the past to surface TLDs doesn’t work at this scale, nor does it handle the vertical nature of nTLDs.
I agree that the entire approach to domain discovery needs to be reconsidered, but this is a big undertaking. In the interim, Registrars should be working to improve their search results. Rightside identified this need early in the nTLD program and started work on our Domain Suggestion Service (DSS). This service supports *all* TLDs in an unbiased manner and does an exceptional job of presenting relevant TLDs, even when using the limited context that a domain query provides.
The Domain Suggestion Service is in use on Name.com and enom – it’s also being implemented by a growing number of other Registrars that share our view that relevancy based results need to play a bigger role in domain search.
We’re looking for partners that are interested in adding highly relevant results to their search capabilities. If you’d like more information, please contact us at DSS-info@rightside.co
Thanks,
Scott Ryder
VP, Channel
Rightside Registry
Thanks, Scott.
I totally understand that in the end right now we need to work and try to improve what we have, and I’d definitely be interested in talking to you “offline” about your service.
The point I was trying to make is however different. My intention was not to complain about the state of “domain search”, but rather to highlight the fact that even the best Registrar interface we have today is inadequate to the task, and this makes it potentially ripe for disruption from someone willing to take a completely new approach to solving this problem.
Maybe I am just dreaming… but I think it’s definitely an interesting topic for discussion 😉
Francesco,
We’re in complete agreement. Domain search is just one component of ‘domain discovery.’ Registrars need to rethink the entire user journey, not just search results themselves. That said, domain search is something that can be improved with a tool like ours or DomainsBot’s in a matter of weeks where as re-architecting domain discovery as a whole is going to take more thinking and planning. Ideally, Registrars should be working on both fronts simultaneously.
Cheers,
Scott
Coincidentally, one of the Product Managers at Rightside Registry published a blog post about search today as well. Enjoy!
http://rightside.news/domain-search-adapting-new-tlds-part-1
Scott Ryder
VP, Channel
Rightside Registry
Spot On!
“My strong belief is that the value of new TLD is not in the simple fact that they are new, but rather in the fact that they offer relevance and a better way to add useful context to an online presence.”
Too many overlook relevance. The cost to market obscure domains year after year is rarely considered.
ICANN’s is not doing enough to advertise,
Registrars do lack effort, the Registries do have unrealistic expectations.
Advertising is not “optional” when introducing a new product.
Registration fees need to be universal. There is no rational explanation for why some domains are required to pay more than any other.
ICANN allowing Registries and Registrars to put themselves before the open market is another step toward balkanizing access by carving up the digital space to work under their own rules.
All three must support *all* TLDs in an unbiased manner.
End users and Investors cannot make informed purchases or long term investments in the new enviroment ICANN is responsible for.
If people don’t know that there is a Thai Restaurant, you can have the menu as large or personal as you want,they won’t find the place. Therefore ICANN and the Registries have to put more effort in promotion and awareness in general.
Very interesting post, and hopefully it provokes some thinking with the players that lag behind in this.
Caro Francesco,
I agree that our industry could profit from a more personalized interaction with its customers and that understanding their “purpose” is key.
I like your ideas of suggesting domains even after users build their site or asking them what they want to achieve. Personally though, I think this would be relevant to some customers but not necessarily to all. As a predictive analytics enthusiast I believe there is much more we can do to engage every customer effectively (even the lazy ones like me).
Allow me to offer a few concrete examples of how our company is bringing personalized marketing to our industry.
During the last 30 days DomainsBot top 10 registrars report a 11% increase in legacy TLD registrations and a 27% increase in new gTLDs due to semantic variations. We help registrants discovering new TLDs by extracting concepts from their queries and moving them to the right of the dot (no pun intended) presenting them relevant, shorter and available domains. I like to believe this works because it’s simple and does not request any particular effort from the client.
We rank the TLDs for each of our registrar or reseller partner according to their conversion ratio (registrations/exposure) so that even minor TLDs have a fair chance to climb to the top of our suggested domains on registrars, resellers and domain monitoring reports as long as they are relevant to the user query and generate registrations. Some new TLDs already outrank some of the legacy ones! You are right, it is the end of the domain search as we “knew” it.
Registrars knowing their customer purchase preferences can also communicate with our API to inform us of particularly interesting TLDs (for instance from their customer shopping history or location).
When a few keywords, location and the purchase history are not enough to match customer purpose to the right vertical, millions of existing websites offer many actionable insights. We are bringing the power of big data to our industry allowing our customer to discover relevant segments of their customer base to target with pro-active marketing campaigns.
Building on your restaurant example. We could send flyers about the new steak house to meat lovers and mention our sushi bar to those chatting about “nigiri toro”. We could show the lager menu to Irish customers and the Barolo to Italians. We could serve dessert to those that always ask for one.
We could serve our customer even if they did not know what they want yet.
Ciao Micio!
As you know, I am a huge fan of the work you guys are doing, and I think you have captured the essence of my article.
I focused on “domain search” because this is what everybody talks about, but I am now thinking that a more accurate title for my piece would have been “it’s the end of the traditional Registrar…”, and time for someone to fully take advantage of AI and predictive technology to help customers find not only the right domain, but also the right services and products to be successful in whatever project they are working on.
Let’s just say for arguments sake that new TLDs are superior to .com domains, it’s still an argument of BETA versus VHS. Nobody wants to be using a system that nobody else is using.
Very interesting post though, and well thought out, however all the King’s men can’t fix Humpty Dumpty.